Advancing Microbial Risk Assessments of Subsurface Water Sources

Abstract

Waterborne pathogens are significant, ubiquitous threats to public health. Thus, microbial water quality evaluations comprise a critical component of the multi-barrier approach to delivering safe drinking water. These assessments underpin the selection, design, and management of drinking water treatment processes. However, the selection of the right combination of tools from an ever-expanding repertoire is essential. This dissertation informs prudent water quality monitoring using existing and emerging microbial tools to accurately characterize microbial water quality and associated risks, especially in sources derived from subsurface environments.

Microbial non-detects cannot be directly construed as indicative of low microbial concentrations due to inherent limitations of sampling and analyzing large volumes of water and imperfect microbial analytical methods. Existing conventions to report microbial non-detects as a measured concentration less than one microorganism within the analyzed volume (the purported method detection limit) were demonstrated to be misleading; handling these values as mathematically "censored" concentrations were shown to result in bias (Chapter 2). Appropriate reporting conventions and statistical approaches were recommended to support the accurate portrayal of microbial non-detects.

The minimization of microbial sampling effort while maintaining adequate precision is a key consideration for monitoring programme design. However, the imprecision of information about concentration associated with small sample sizes is seldom explicitly quantified. Using simulated protozoan monitoring data, the attainable precision of the estimated mean protozoan concentration under different hypothetical sampling scenarios was evaluated. A framework was developed to quantify precision and contrast the relative merits of additional sample collection for protozoan enumeration from the source versus characterization of method analytical recovery (Chapter 3).

In microbial groundwater quality evaluations, the need for sufficient well purging to obtain representative samples of microorganisms suspended in aquifer pore water without artefacts attributable to well-related biofilms has been widely recognized. Adenosine triphosphate (ATP) was therefore evaluated as a rapid, field-ready biochemical indicator of microbial water quality changes (Chapter 4). Supported by concurrently measured microbial water quality parameters, ATP measurements exhibited phenomena reflective of time-limited (bio)particle transport behaviour. Microbial groundwater quality assessments must therefore be designed using approaches that are necessarily different from those used to describe dissolved solute transport behaviour. A subsequent focused investigation of one biomolecular tool—bacterial community analysis based on 16S rRNA gene amplicon sequencing (Chapter 5) demonstrated its utility for identifying fecal source and surface connectivity indicators (e.g., cyanobacteria). Factors contributing to bacterial community variations were examined. Collectively, these assessments indicated that an appropriate suite of microbiological tools can be concurrently utilized to overcome the challenges of spatial heterogeneity and dynamic hydrogeological conditions to meaningfully characterize microbial water quality at the aquifer scale.

In this dissertation, existing and emerging microbial tools to support groundwater vulnerability assessments to fecal pathogen intrusion were critically examined. The persistent need to consider the "fit-for-purpose" ability of these various tools to support microbial water quality evaluation is emphasized. Moreover, this dissertation underscores the complementary use of these microbial tools to inform groundwater vulnerability assessments; no single tool will entirely capture the elusive, multi-faceted nature of subsurface microbial water quality.